A potentially good site for a broad band seismic station is on hard-rock and far away from cultural noise. However, in remote places, such as Botswana, most places have soft and sandy soil and theft or vandalism of the equipment is a great concern. We choose to install most stations on terains of primary schools which have 24/7 garding and are well distributed across the country and taking the higher ambient noice for grand.
Most areas in Botswana have sandy soil and few of them have hard rock or clay. At the sandy locations we use our "Flow-Tek water tank setup" as illustrated in the figure below. We dig a steep pit that will fit to the size of a 2500 Litre water tank (Flow-Tek Botswana) and pour a 15 cm layer of concrete at the bottom of the pit. Te tank, from which the middle part of the bottom was cut out, is placed in the pit and extra concrete is poured through the man hole at the top. Finally we end up with a bottom thickness of 25 cm or more. To avoid signals due to expansion problems, no iron is used. The concrete will stick to the underground and by covering the tank bottom slab, a water tight construction is created. On top and bottom the tank tube wholes are used for air ventilation and covered with gauze against animals. To avoid people climing on the roof
The sensor is placed in a bottom less bucket, directly on the concrete bottom. After adjusting the sensor, the bucket is filled with sand bags for extra temperature isolation and can easy be removed in case of a sensor problem.
To keep sun and rain away, a sloping roof is placed over the tank. The tank man hole lid has screw-thread and therefore easy to lock with a padlock.
The tank inside working space is 140 cm in diameter and more than 150 cm in height. This space is enough for all intruments and battery's and still comfortable to work in.
In those cases where the ground is to hard to get in, we build a small shelter from stone. Hollow stones and double roof are used to have at least some temperature isolation.
The "A team"
Without the effort of our DGS colleague in Botswana, it would not be possible to set up this seismic network.
The seismometer should be oriented as close as possible along the geographical NS/EW direction. A magnetic compass measures the direction to the magnetic pole, but has to be corrected for the local declination. At station locations with metal in the construction a magnetic compass is impossible to use and therefore we use a Gyro compass in combination with a magnetic compass. First we measure the magnetic positions at least at two different places to be sure that the presence of metal in the ground did not disturb the measurement. Then the gyro is exactly positions along the compass and adjusted for declination. Now the gyro points exactly to the geographical NS/EW direction.
To be independent from local the local electricity net, all sites are powered by solar energy. The total power consumption of a NARS station is only 15 Watt therefore a panel of 140 Watt is installed in combination with 2 x 105 Amp Hour batteries. When fully charged, it will keep the station running for at least 5 days in case of overcast. An intelligent charge controller regulates the power and takes care for battery overcharging. The data logger is connected directly on the batterys and not at the controller load. Only the GSM modem is connected at that terminal. Depending on the battery charge level, the data logger power monitor will shut down or reboot the data logger system. This will avoid an uncontrolled power switch off by the charge controller what could lead to a data logger disk crash. The solar panels are mount on poles made of old drill pipes coming from DGS. The panel is mounted in a slot frame and closed with a padlock. On top of the frame we mount the GPS modem. Cables are going inside the drill pipe and in the ground to the Flow-Tek tank with instruments.